Latest / State of the AI Union / Ben Orthlieb: Can your VC do this with AI?
Transcript
- 0:04All right, hello and welcome to this week's episode of the State
- 0:07of the EI Union. I'm your host, Laura, and joined
- 0:10with me today is Ben Orslade of Blue Moon Ventures.
- 0:14So, Ben, would you like to introduce yourself and also
- 0:16your? Company, of course.
- 0:20So Ben Orslade originally French built Blue Moon, which we're
- 0:24going to talk a little bit about.
- 0:26But Blue Moon came from the idea that you could probably compete
- 0:30in venture by using technology, which has led to a lot of
- 0:35interesting discoveries along the way.
- 0:37Turns out it works. Not only are we investors in
- 0:41Devrev, but a number of other companies.
- 0:43And so our fund so far has been beating 99% of VCs, which is
- 0:49will not surprise most founders. Good tech plus a good process
- 0:51actually works most of the time. That's the whirlwind.
- 0:56Before that, I spent the last 10 years of my career at LinkedIn
- 1:01where I was leading M&A, investments and corporate
- 1:05strategy. Awesome.
- 1:07Thanks. Ben.
- 1:08So you briefly touched on it and you said like how can we use AI,
- 1:11help us, you know, build a better VC firm.
- 1:14So tell us about how you're doing that right now.
- 1:16How are you using AI to help you?
- 1:18It's very interesting. And I think the, the, the key
- 1:20for us is we're using AI to do what most VCs do.
- 1:25We're not reinventing the process or reventing what
- 1:27investing is. We're just doing it faster,
- 1:31better, cheaper in some ways with AI.
- 1:35And what I want to make clear is it, it's not making decisions
- 1:40for us. We're not algorithmic trading
- 1:42because this is still a private market.
- 1:44And so you actually need to convince people to take your
- 1:46money. And a lot of the signals in
- 1:49venture in the core of our decision making is on the on the
- 1:51founders themselves. And that is a very deep personal
- 1:56conversation, human to human, if you will.
- 1:58But how we use it is really across the entire workflow.
- 2:05So maybe I'll start with Step 2, which is screening because it's,
- 2:09it's the important piece to maybe crack first.
- 2:14We use unsupervised machine learning to screen founding
- 2:17teams based on their background so effectively professional
- 2:23experiences, academic experiences, personalities
- 2:27inferred from what they say. If we can find enough words
- 2:31either written or spoken online, this helps us spend time cut
- 2:37through the noise, right? If if you sourcing is always
- 2:41problematic because you have to screen and screening usually
- 2:44takes a lot of manual time, a lot of associate times in
- 2:48traditional venture, you know the legacy funds, they usually
- 2:51do their downloads of harmonic on Monday, send spreadsheets to
- 2:56associates and they have to somewhat find the time to go
- 2:59through those spreadsheets and manually sort of score people.
- 3:04Our approach is 1 where we say, look, there are patterns.
- 3:08If you actually abstract what the the algorithm does, you find
- 3:147 or 8 pockets of types of founding teams.
- 3:17We score the founding team that tends to be more successful.
- 3:21And so of course there's way more than just screening that we
- 3:24do. But the the idea is you need to
- 3:27cut through the noise because no matter what you do, you have too
- 3:30much sourcing and the ability to basically trust an algorithm
- 3:34that will tell you a better pool of people, more concentrated
- 3:39people, pool of people to spend time with because they are more
- 3:41likely to succeed in venture. That's what the algo does.
- 3:46Now. The interesting thing is once
- 3:48you've solved to some degree Step 2, then you can go back to
- 3:53step one on sourcing because now we can cut through the noise and
- 3:57spend no time on it. And so over the years, we've
- 4:00seriously amped our ability to source founders and founding
- 4:06teams. To give you a sense, last year
- 4:09we sourced 1100 companies and teams that were already
- 4:18qualified as being CB2B North America, which at the end of the
- 4:20day, what we do is plain vanilla is really how we do it that's
- 4:25different and sort of innovative.
- 4:28And so sourcing and screening at this point is fully automated.
- 4:31We call the system Agatha. I don't know if you have seen or
- 4:36read Minority Report, but in Minority Report there are three
- 4:41people that can see into the future.
- 4:43The best of them is Agatha. And So what we say, you know,
- 4:47we're not marketing pros apparently, but what we say is
- 4:49Agatha sees into the future of founders.
- 4:52And so Agatha helps us to, at this point, abstract away
- 4:56sourcing and screening. We keep adding more ideas for
- 4:59how we source because at this point we just want to see really
- 5:03everyone. And then Agatha without any
- 5:08other signals effectively already improves your chances of
- 5:12success here, measured as graduation by about 50% compared
- 5:17to randomly picking without any other signals.
- 5:20Amazing, amazing. And you'd you said 11,000
- 5:23companies you you source and and you guys are just two people,
- 5:26right? Like you don't have any other.
- 5:28That's right, this sounds amazing. 11,000 companies with
- 5:32two people. Last year 11,569 to be precise,
- 5:38What's interesting is you know, we, we sourced about 1200 from
- 5:42our network, meaning the sort of classic approach and it's
- 5:46already 1200 is already pretty good deal flow for an emerging
- 5:50manager. But we basically 10X that by
- 5:54using data and all sorts of different techniques and
- 5:57sourcing pipelines to really sort of literally 10X that.
- 6:02That's amazing. So if if we're if we, if we sort
- 6:06of move towards this world, world where sourcing and
- 6:08automation, you know, become really become really automated
- 6:14and there's no need for really sort of like the VC analyst role
- 6:19anymore where where somebody has to like score the score the
- 6:23founding team. Wait, where do you think that
- 6:28where do you think that that learning is going to come from?
- 6:31You know, how will how will the expertise sort of develop?
- 6:35Yeah. It's actually, I mean to be fair
- 6:37to analysts that they do more than just core people on
- 6:40spreadsheets. I think what was interesting is
- 6:44actually our next step, which is where it's a little bit more
- 6:47difficult maybe for associates role is once we found and scored
- 6:54people, the next step is to analyze what the companies do
- 6:57first before I meet them and then of course, based on the
- 7:00conversation. And there we've built an AI
- 7:02associate that we call Blue that gives me that detailed company
- 7:08analysis. It tells me essential questions
- 7:11to ask the founders in a first meeting, follow up questions
- 7:14after a meeting and it does market research and analysis.
- 7:18What's interesting though, is that piece is work that
- 7:22associates cannot do today because they don't have time.
- 7:26And so and it's, you know, some people are doing deep research
- 7:30right now. It's, it's what we do is a very
- 7:32highly specialized rag with a very deep sort of knowledge base
- 7:37because, you know, a prompting ChatGPT doesn't cut it for
- 7:41smaller companies, but I think it's now more enabled us to do
- 7:48way more that most companies cannot do anyways, even if they
- 7:52have lots of associated manpower, at least for the
- 7:56screening part, we can just do more.
- 7:59And associates frankly don't particularly enjoy the than the
- 8:03hours on spreadsheets. And then the preparation, to be
- 8:08honest, like most VCs do not have time to prepare meetings,
- 8:11at least the way to the level that we do.
- 8:14In fact, we are now sharing that with a few VCs that are helping
- 8:17using this to either screen themselves or to do meeting
- 8:21preps. The question do your do your
- 8:23question like associates are still going to be very important
- 8:26in meetings once you engage with the founders, but it really
- 8:31everything beforehand can be automated.
- 8:34And honestly, I think that's where the learning to your to
- 8:37your point or to your question, the learning comes from the
- 8:41iterations on meeting founders, debriefing with GPS, not so much
- 8:46from almost the grunt work that happens before the grunt work
- 8:50has to be done. It was done manually.
- 8:52It's now done, at least for us, with with AI and tools.
- 8:57Or even in like retraining the agent and giving it more
- 9:00information. Right.
- 9:01Those are the things that they can learn from.
- 9:03Yeah. Awesome.
- 9:04OK. Where where it gets dicey is
- 9:06once the analysis we get after a first meeting, both on the
- 9:09company and on the founders, is becoming very, very strong.
- 9:15And let's talk about that. So like now that you are,
- 9:20you've, you've, you've met a founder and they've already been
- 9:22scored. So they're, they're all good on
- 9:23the first pass. You know, you're, you're meeting
- 9:25with them. What do you think or how, how do
- 9:28you think, how do you think founders should prepare for
- 9:33those conversations, given that they've already been pre
- 9:36screened and this pre screening rate is like very, very high,
- 9:40very accurate. It isn't, you know, they don't
- 9:43have to change your mind on some like non objective, some
- 9:48subjective qualification, that kind of thing.
- 9:50Yeah, yeah, I mean, they, they need to be on our process.
- 9:56This is just step one. I need, I, I only need people
- 10:00with a certain score. Yes, I will only meet people
- 10:04that I've scored and, and so but, but, but at that point, you
- 10:08know, whatever your score is, I at that point do not care
- 10:11anymore. Yes, because it's just a
- 10:13screening and it's imperfect. Of course there's type 1, type 2
- 10:18error, but I have learned to overtime trust the process and
- 10:23just know that you know, ultimately the main constraints
- 10:26on venture is time to meet with founders.
- 10:29And so I do not have the luxury of being able to meet 11,569
- 10:35founders last year. I can probably meet between 3
- 10:37and 400 a year and that's how I sort of put my screen.
- 10:44After that is human assisted by AI for decision making, but it
- 10:50is still human decision. But let, OK, let's talk about
- 10:54before they get selected, since since we know that the founders
- 10:57are being, you know, pre screened and it's pattern
- 11:00recognition, as you said, what if the founders are not part of
- 11:04the pattern? I I don't maybe that's part of
- 11:05the pattern itself, right. How how can they prepare?
- 11:09How can they prepare to get noticed also ahead of time?
- 11:13Yeah. There's two elements.
- 11:15One is what's been fascinating in unbundling our algo, if you
- 11:22will, is of course there is the pattern that most VCs kind of
- 11:26screen for by default, right. You went to Stanford or MIT, you
- 11:31went to Google or Facebook for a couple years, you've already
- 11:33done AYC company. Those people tend to score well,
- 11:38but we've found that there are many pockets that get mostly
- 11:43ignored by VCs that still get picked up by an algorithm, which
- 11:48actually is a good thing because it means now you don't
- 11:50necessarily have to have the just the one flavor of founders
- 11:53that VCs, legacy VCs tend to like.
- 11:55You can actually be different and succeed.
- 11:58So a critical example or a clear example of this is Mercor.
- 12:03We invested at Mercor in the seed round alongside General
- 12:08Catalysts. And so of course a few funds
- 12:12like them, but most funds actually screen them out because
- 12:16Marcor is 3 first time founders dropouts, which dropouts can be
- 12:23interesting in B to C, not so in B to B.
- 12:25It's not part of the usual sort of profile, if you will, The so
- 12:31they didn't fit the the usual profile yet they scored very
- 12:36well on the on our evaluation. So we engaged with them.
- 12:40Now, how to get noticed? I think it's not so much for US
- 12:45founders because they tend to be quite prolific on LinkedIn apps,
- 12:49etcetera, but I think being sharing information out there
- 12:52about you, we capture stealth founders.
- 12:57Stealth used to be very hot topic in VC, not so much, it
- 13:02turns out, because stealth without other information I'd
- 13:06seen, which is what we do. And so not exactly your first
- 13:08check tends to be a lot of noisy evaluations and and so random
- 13:13meetings. And again, the critical thing is
- 13:15time. So it's cool to be in stealth.
- 13:18It's a good way to get detected maybe.
- 13:20But and we find about 510% of our the founders we capture are
- 13:28in stealth. But if you want to have the
- 13:33right VC conversation, it's it's better to be a little bit more
- 13:35forthcoming, I would say, yeah. Yeah, yeah, yeah.
- 13:37And it's, it's really nice to to hear.
- 13:40I think that, you know, even though your scoring is based on,
- 13:45on patterns that, that it is picking up those, those
- 13:48opportunities that are not necessarily a part of the, the
- 13:52legacy, you know, the, the legacy pattern and what, what
- 13:56they're what they're looking for, right.
- 13:57I think that's the thing that people are afraid of that
- 14:00they're not. And one key output is diversity.
- 14:05And so it is not diversity by design, right.
- 14:09We, we didn't force it into the algo, but it turns out if you
- 14:12look out a portfolio, which is effectively what we do, you end
- 14:15up with a portfolio that's 10X more diverse than the average
- 14:19inventor. So 25% women founders, 40% women
- 14:23Co founders and and sort of true across all statistics.
- 14:27And it's interesting because this is the outcome, not the
- 14:30inputs. Yeah, yeah.
- 14:32So it tells you you you look at different criteria.
- 14:35Right, right, right. So, you know, when we we talked
- 14:37in our prep that about how you show your your preps to your
- 14:42founders, you share the information in the analysis with
- 14:45them. And I saw it and it was very
- 14:47impressive, right, that that these founders are coming to
- 14:52have a conversation with you. Knowing that you already have
- 14:56this level of technical expertise.
- 14:57How do you think that changes that conversation around their
- 15:02around showing their product or how do you think it should
- 15:05actually change, you know, if if you're not sort of getting the
- 15:09conversations that you want in in, in those meetings?
- 15:12Yeah. For us, it's a it, it was a bit
- 15:15of a game changer when we decided to test sharing it
- 15:23because one of the key metrics in our flow is, you know, we
- 15:27find we find you, Laura and Agatha.
- 15:30Agatha thinks you're amazing. Now most of the time I don't
- 15:34know you, so I need to get in touch.
- 15:37And So what we do is I would send you an e-mail saying, Laura
- 15:41Agatha, my AI found you and thinks you're a great founder.
- 15:46And I've already been doing work on what you're doing, so I'm
- 15:50prepared. And so here's a link to my AI
- 15:52associate analysis. Here are some very tailored
- 15:55questions. And the more, you know, when we
- 15:57started, we were a little bit shy about sharing all those
- 16:01things, right? And it turns out founders love
- 16:04it. And it makes sense in
- 16:05retrospect. The idea becomes from the touch,
- 16:08first touch points, we're not traditional VCs.
- 16:12We are tech founders. We are tech founders in the VC
- 16:16space, but we're effectively more of a Fintech.
- 16:19And so the number of times I've had founders tell me a version
- 16:23of I'm building all day long with AI, but it's the first time
- 16:27I'm seeing AVC that's building with AII.
- 16:30Really want to talk to you. It, it, it's been really sort of
- 16:33a bit of a game changer in terms of the ability to without asking
- 16:38for intros etcetera, just get, get through and have direct
- 16:40conversations. So that that's been quite
- 16:45important. Some things we found over time
- 16:48as we were building it, that really changes the dynamic.
- 16:51Of course, we're also a small check.
- 16:53Our strategy is, you know, 250K check as part of a bigger round.
- 16:57So the, the sort of stakes are different, but our ability to
- 17:03find you and talk to you is 75%, which most outbound VCs from top
- 17:09phones is sort of in the 30% range.
- 17:12So that really tells you not only do we get informational
- 17:16advantage because now I show up in a meeting very prepared.
- 17:19So I, I have real questions to ask, but it also actually has a
- 17:23very clear impact on our winning ability by effectively
- 17:30positioning us as a different sort of as more founder alike.
- 17:36Yeah. And, and I think we, you know,
- 17:37we also talked about this where it's like, hey, if if I am an AI
- 17:43founder and you're coming to me with, with, with that kind of
- 17:46expertise, that's more likely to be an opportunity for me to want
- 17:50to work with you, you know, versus, you know, maybe even
- 17:54like a big name. On that front, because it was
- 17:59part of your question also, I've had founders that were like, oh,
- 18:01you guys have built versions of what we're trying to do, of
- 18:05course for a different market or things like that.
- 18:07And so that creates sort of very interesting conversations
- 18:11because the founders understand that they can go one left, one
- 18:15click deeper from the get go on what they're doing and how
- 18:18they're approaching it because they understand from what we
- 18:22send them the kind of things we do.
- 18:25And you know, we, we, we, you briefly touched this earlier
- 18:29where, you know, you said you had a really high, high win
- 18:32rate. And now that founders are able
- 18:36to find investment opportunities like Blue Moon where the
- 18:41founding team has experience with AI and you know, they're
- 18:44doing a, a lot of these things. How, how do you think it's just
- 18:47in general that has changed the the outlook of the investment
- 18:52market and how people may be shifting away from these brand
- 18:55names, you know brand name VC firms etcetera?
- 18:59You think brand names still matter because venture is
- 19:07partially a game of branding and momentum.
- 19:13And so having a good VC on your cap table at a certain point
- 19:18still is a signal to the market that helps you recruit, that
- 19:22helps you reassure your customers that yes, you're early
- 19:24stage, but look at the investors are backing us now.
- 19:30I that said, founders are a lot more careful or skeptic in their
- 19:35choices because they understand that a lot of VCs haven't coded
- 19:42in many years. And the fact is only 14% of GPS
- 19:46are engineers by background. And it's not a Ding they've been
- 19:52doing something else. But the most part haven't been
- 19:54coding for the last 20 years. And so I would say the
- 19:58opportunity is for funds like us to carve a niche because all the
- 20:08sudden you're more attractive to, to to founders.
- 20:10That said, it's, it's also part of why we're a small check and I
- 20:14don't compete to lead you. You still need a very strong
- 20:20brand name VC to help you out. Or at least to share their brain
- 20:25with you really is sort of what it comes down to.
- 20:28But you know what? One Part 1 fun.
- 20:31And one person that's been very good at sort of carving out that
- 20:33niche is Thomas at Theory Ventures, right?
- 20:37He's been writing about tech pretty much every day for 10
- 20:40years. They have real technology to
- 20:45help them do venture and that helps him compete and carve A
- 20:49niche for himself or a good niche for himself in the market.
- 20:53Now he leads rounds. And so he needs a lot more of
- 20:56that, which he does have given his background, but he he, he
- 20:59needs a very high level of credibility to win against some
- 21:03of the bigger names. But it is surprising to us that
- 21:07very few VCs have evolved from the traditional model of
- 21:11branding by sort of podcast and sort of blog post without really
- 21:16going down the path of competing on tech.
- 21:19So there's only a handful of funds that are competing on
- 21:21tech. OK.
- 21:23And I heard that you that Harvard, Harvard Business School
- 21:28is doing a case study on PD on tech thing.
- 21:31Tell us a little bit about that. Yes, it's been a, it's been a
- 21:35fun ride with the the team at Harvard, but we have been the
- 21:41object of a case study, if you will.
- 21:43The case study is actually kicking off next month.
- 21:45So we'll be in Boston for the first sort of student discussion
- 21:49on the case. But the idea is, is an
- 21:53interesting one and one that we asked ourselves.
- 21:55The question is really, OK, you've developed all these tech
- 21:58for yourself, right? It is very tailored to our
- 22:00process. Now, should you share, should
- 22:05you sell some of these elements to with other VCs?
- 22:11Do you become a product company? Do you become, do you stay a
- 22:15venture firm? Are you somewhere in between
- 22:17where maybe you share technology with LP's and VC friends, which
- 22:23is the path we're on right now. But yes, that's been a
- 22:27fascinating journey in part because going through the
- 22:29process was also forced reflection on our part.
- 22:33So that was, that's quite an interesting, an interesting
- 22:36outcome. But yeah, it it, it also does
- 22:39tell you that while every other asset class uses a lot of
- 22:43technology and data, venture is very much in the early stages.
- 22:49And so a firm like us, small, relatively unknown brand, still
- 22:54makes it to Harvard as a case study, as something novel and
- 22:57interesting. Yeah, that's very interesting,
- 22:59isn't it? Because all the money is going
- 23:01into tech and it's still, you know, a lower tech industry.
- 23:07OK, cool. So we know all now that
- 23:11basically every founder out there that's raising money is
- 23:13probably doing something AI related.
- 23:15So, you know, it's mostly, you know, help me raise, help me,
- 23:19you know, fund my, my AI idea here.
- 23:22What are what are some things that you would encourage a
- 23:25founder to think about to really show their value add and what
- 23:32should they be asking their potential investors?
- 23:36On the first one, it it, it still is, yes, AI is the
- 23:41technology you have to use at this point, but it really comes
- 23:44down to classic almost Harvard like strategy, like how is what
- 23:49you're doing differentiated hard and given that technology is
- 23:57going to be easier and faster. So hard technology becomes way
- 24:01harder. So if you have an advantage
- 24:03there, it's probably very strong, but the middle part of
- 24:08technology that that was a lot of software is actually become
- 24:11super easy. So what makes your company still
- 24:15a potential winner in a world where a lot of it can be copied
- 24:20or replicated by other companies?
- 24:23And so it comes down to positioning key product
- 24:25insights. I think the value of domain
- 24:29knowledge has actually gone up tremendously because now most of
- 24:33tech isn't the limits so on what people are doing.
- 24:37And so it becomes down to, it comes down to really domain
- 24:40knowledge and understanding, deep understanding of your
- 24:43customers. Yeah, and.
- 24:45Sorry, Ben, on that, is it, You know, I think what you're saying
- 24:50is that, that, that the, the investment itself should not,
- 24:56should not really change because we're, you've always been
- 24:58investing in solving business problems.
- 25:02And it's just that today, you know, we're using AI and you
- 25:04have to be using AI to do that. So do you think on the messaging
- 25:08side, on the pitch side, if I was a founder and I was to to
- 25:10VC, should I start with, here's the problem that we're doing and
- 25:14we're solving for it and by the way, we're using AI to do it or
- 25:17should we start with we are using AI to solve this business
- 25:21problem? I think there's two paths.
- 25:24If you're doing deep model level, deep sort of AI
- 25:29technology, you should say I'm doing deep AI and this is what
- 25:34I'm doing and I'm solving a real infrastructure, a real technical
- 25:38problem. If you are more of a software
- 25:42solution with AI, it's sort of it's if you're not do if you're
- 25:48not using AI is mostly like that would actually be mostly a
- 25:51reason for questioning, but it's not the oh, I'm doing vertical
- 25:54AI and real estate. Great.
- 25:57I mean, so fine, like everybody like there's a lot of companies
- 26:00that are doing that. So yeah, I I think build it.
- 26:04It depends. Then the application building
- 26:06with AII don't I haven't seen a start-ups in week that hasn't
- 26:10told me they're building with AI.
- 26:11Of course, everybody's building with AI.
- 26:14It's not the same as being a deep AI company if you want to
- 26:17use this term. I can't.
- 26:19You have the. I think you're going to answer
- 26:20the second question, which is what what founders should be
- 26:23asking their VCs. It's a very good one.
- 26:27There's the usual what can you help me with?
- 26:30I don't know if it's actually that helpful.
- 26:33I think you should. The what you should be asking
- 26:37their VCs is probably make me intros.
- 26:43I found that people that of course you need to be a little
- 26:47bit more chaste in that my tier ones, but asking VCs to prove to
- 26:53you along the process that they can add value is actually better
- 26:55than any warning. So one way we do that for
- 26:58example is we're small, we invest as part of larger rounds.
- 27:04We make lots of intros to bigger VCs that can lead during our
- 27:08process. Last year, in fact, we made 278
- 27:11intros and again, or small fund, we don't have a brand.
- 27:17You take the meeting because you're like, oh, blue moon is
- 27:20doing tech That's cool. I want to learn more.
- 27:23But the fact is through your process, I'm very likely that I
- 27:26will give you a handful of intros to tier one funds and
- 27:30that makes my value tangible to you.
- 27:34There's a whole bunch of other things that I do after
- 27:36investments, but at least, you know, I'm, I, I'm putting in
- 27:39some work upfront. And so it, it's really, I think
- 27:41this question like what what can people do for you up front is 1
- 27:47and then the other is I would ask VCs or honestly previous
- 27:54portfolio companies, how the VCs and the brand has helped them
- 27:59with getting more customers or better hires.
- 28:03So really like dig, dig down, you know, can they help you get
- 28:08more introductions to VCs customers and you don't ask them
- 28:13to kind of prove it along the way is what I'm.
- 28:15Here, yeah, the reality is some VCs add value beyond that.
- 28:20But at the early stages where I'm at, the real question is,
- 28:24does the brand and or what they do help you grow faster and grow
- 28:29faster because you secure more sales, you secure more or better
- 28:33hires? Yeah, awesome.
- 28:35OK, So Ben, let's learn a little bit more about you.
- 28:38You transitioned you, you mentioned that you were from
- 28:40France and then you transitioned to the V area.
- 28:43Tell us a little bit about what brought you here.
- 28:46Oh. I've always wanted to come work
- 28:51in the USI grew up, my dad was a commercial pilot, and so I have
- 28:57been to Los Angeles an inordinate amount of time for a
- 29:01French teenager. And so I always had this idea in
- 29:04the back of my mind. And so I find my way, found my
- 29:06way to San Francisco first, working on for a French Bay on a
- 29:12rotation program that clearly was leading me here.
- 29:17OK, OK. So that's how I landed here.
- 29:20Things evolved. I did an MBA and then sort of
- 29:24permanently came back to the Bay Area.
- 29:26OK, fantastic. And what's 1 French food that
- 29:29you just absolutely cannot find over here?
- 29:32Absolutely cannot find I it is very hard to find, but still
- 29:39exist. Both.
- 29:41Bourguignon is like a a Stew, but there's a secret ingredient
- 29:46to make it really good, which is you need to put a a a piece of
- 29:49chocolate in the Stew at the last minute.
- 29:53Very few people know that, so there's only one or two
- 29:55restaurant I think in the Bay Area that get it right.
- 30:00Which ones are those? There's court West in in the
- 30:06Marina and then in Sausalito, this place called Suzette, they
- 30:10have it on a special in a rotation.
- 30:13Awesome. OK.
- 30:14And last question here, what, what do you think is one thing
- 30:17that Americans can learn from Europe in technology and
- 30:24interacting with let's say VCs, etcetera?
- 30:28It's funny because usually it's the other way around.
- 30:30Like I think Europe still has a lot of to learn about the
- 30:33venture ecosystem and, and, and building start-ups.
- 30:36One thing to learn, one thing that's been interesting is
- 30:40there's a lot of great AI engineers and researchers coming
- 30:45from Europe and France in particular.
- 30:49And my theory on this is, and I sort of, I, I, I was doing
- 30:54machine learning as part of my engineering study way back.
- 30:58It was a space that at the time wasn't seen as commercial.
- 31:01And I think that's why on a relative basis, the US sort of
- 31:04deprioritized it compared to Europe that kept doing research
- 31:07on it. So maybe the learning was be a
- 31:10little bit more patient on on seeing potential economic
- 31:14outcomes when interesting tech comes along.
- 31:17Well, that or that there's a there's a lot of talent there
- 31:19that might be untapped. That is for sure.
- 31:22Right. Yeah, OK, awesome.
- 31:24Ben, thanks so much for being on the show.
- 31:26Is there anything else that you'd like to leave our
- 31:27listeners with? I think, I mean, people
- 31:31listening to you and the Devra Devra team probably know this,
- 31:36but experimenting with AI opens up so many ideas and so many
- 31:40sort of new avenues for how to reinvent the process.
- 31:44At the end of the day, what we're seeing is what we do is
- 31:46still unique 5-6 years in, not because of a lack of tech, but
- 31:50because of culture. And so most people, most
- 31:54companies are wary to change their processes, which it makes
- 31:57total sense. However, that's where disruption
- 32:00comes from. And so my small advice is
- 32:04experiment and play a little bit and you'll you'll get down the
- 32:08rabbit hole pretty quickly. I love it.
- 32:11Let's go, let's go experiment. Exactly.
- 32:13Thanks so much, Ben. And to all listeners, if you
- 32:15like this episode, please drop us a comment, leave us a like
- 32:18and subscribe. And until then, we'll see you
- 32:20next time. Bye, Laura.
- 32:22Bye bye. This episode was brought to you
- 32:25by Devrev for Startups. If you're a founder building
- 32:28with AI and have a story worth sharing, would love to hear from
- 32:31you at Startups at Devrev dot AI.
- 32:34Don't forget to explore the Devrev for Startups program on
- 32:37devrev dot AI slash startups where you can get up to 12
- 32:41months of credits and much more to experience true team
- 32:45intelligence with computer your AI teammate.